All right. Good afternoon, everyone. Thanks for joining us at the Morgan Stanley Global Healthcare Conference. I'm Mike Ulz, one of the biotech analysts here, and it's my pleasure to introduce René Russo, CEO of Xilio Therapeutics. Just a reminder, the format for today is a fireside chat, so if anyone has a question from the audience, please feel free to raise your hand, and we'll be sure to address your question. Before we get started, I just need to read a quick disclosure. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. With that, thanks, René, for joining us. I thought, maybe we could just start with a macro question.
If we look across the oncology space in recent years, one of the areas of promise has been immuno-oncology, but there's also some risk associated with that. You know, in your view, why is it important to continue to invest in this area and continue to explore-
Yeah.
Opportunities?
Yeah. So I think for us, the way we look at, the promise of IO is that still very much the biology, and the science suggests that the best way to really achieve broad applicability across a wide range of tumors and really focus on what we all want, right, is those long-term durable responses and cures, really requires modulation of the immune system, and we can't give up on looking for those long-term durable responses and cures, right? That is the promise of IO, and it has broad applicability. I think some of the challenges, especially, with new targets, I think sometimes the expectations get a little bit ahead of the clinical data in IO because IO takes a bit more patience, you know, in terms of waiting for those responses because it is different than targeted therapies, for example, or direct cytotoxics.
But I think that focus, right, and that promise is still, is still there. And so one of the ways we're trying to overcome some of the challenges with IO is by focusing all of that power of the immune system just in the tumor, right? And allowing us to get to higher doses, high concentrations, very potent molecules, without really implicating the immune system in the healthy part, healthy parts of the body and healthy tissues.
Gotcha. Makes sense. And maybe you're starting to talk about your platform, so maybe just expand on that a little bit, you know, and talk about what's the key hurdle and how do you sort of get past that or improve that-
Yeah.
Part of the therapy?
Yeah. So our thesis has been, right, if you can harness the immune system, really in the tumor where you want it, and keep it quiescent or quiet in the rest of the body, you know, that would really be the key to unlocking real potential in IO. And the way our technology works is we apply a masking domain that turns off these very potent molecules, so they have no activity outside of the tumor, and then we utilize the dysregulation of solid tumor MMPs or proteases that are part of the normal biology of tumors. That's how they invade, grow, metastasize. We utilize that to then cleave off the mask and turn the molecule on only in the tumor, and this is how our approach has been evolving.
And now, over time, with our platform, we started with in vitro data. We moved to in vivo. We looked at over 1,000 ex vivo human tumor samples, and we saw the molecules activating. But now, for the first time, in fact, just as late as last week, we've now demonstrated in two tumor biopsies in our CTLA-4 program that we're seeing over 70% activation of the molecule in human tumor samples. We saw this in a melanoma lesion and in a hepatic metastasis in a colorectal cancer patient. So we're just seeing that now translate in the clinic, as well as also looking at tumor-selective PD, which we've seen in 4 biopsies from the IL-2 study.
Yeah, so making progress there. Very exciting. Maybe if you just focus on 101, your CTLA program, maybe you can highlight some prior data there, in combination and sort of the path forward, in combination with Roche and why you decided to go that route.
Yeah. So our CTLA-4 program is just wrapping up now phase I-A. So this is a very potent Fc-enhanced tumor-selective anti-CTLA-4. We had a surprising PR that we saw in our phase I data, which was unexpected, frankly, for a CTLA-4 in a phase I trial. This was a Stage IV non-small cell lung cancer patient with liver mets. And we were very pleased to see for this patient that after the first cycle, the liver mets were undetectable on the first CT, and then that patient also had a reduction in the primary lung lesion and with a confirmed PR. Now we're out 36 weeks, this continues to be a durable response. The liver mets remain undetectable.
And I think importantly, the tumor selectivity for this patient really showed up in that their only treatment-related adverse event to date has been grade 1 fatigue, so really sort of unexpected. This is also PD-L1 negative patient. So for a whole lot of reasons, there was a lot of enthusiasm about that response in the phase I. And that response is what led us to this collaboration now with Roche, where we're looking at our anti-CTLA-4 XTX101 in combination with atezo, and we're moving into MSS-CRC.
We'll talk a little bit more about that, but one of the benefits of having the data that we have is that MSS colorectal cancer patients, the large majority, have liver metastases, and are often excluded from IO trials, yet our response in that one patient showed that effect in liver mets. So we will be including patients with liver mets in that trial.
G otcha. And maybe you can talk a little bit more about the design of that study in terms of dosing, and maybe the starting dose as well.
Yeah, so the way we'll start, so first we'll do a combination to determine safety. So, we will take the dose of CTLA-4 down from our RP2D. So the RP2D was 150 Q6 weeks. We'll likely start at 75, see how that goes in combination with atezo at 1200 mg, and then, depending on the tolerability, can go up or down from there. So we would anticipate likely, if all goes well, two dose levels, and then go right into the phase II. And the phase II here we think is pretty unique because historically, the combination, or sorry, the contribution from PD-1s or PD-L1s as monotherapy for MSS-CRC has a less than 5% efficacy rate. So for us, this is, you know, a path to really look at the combination, without having to do a randomized trial-
Yep.
Versus the single arm. We think based on, you know, statistical analysis, we can get to, you know, a reasonable answer with around 40 patients in that phase II.
So that 5%, that's ORR for sort of the expected outcome?
Yes. If you kind of look at all the PD-1, PD-L1 data together for MSS-CRC, it's under 5%.
Yep. And just talk about the reason for... You have a recommended phase II dose, but to start, you're kind of stepping that down. Is there some concern about overlapping toxicities, or is it just, you know, trying to figure out what's the lowest effective dose in a combo? 'Cause presumably you'd have some additive effects.
Yeah, I think it's primarily first to ensure that we don't trip into any safety signals by starting with a lower dose. So we know the 150 Q6 week monotherapy for us was very well tolerated, no grade four or five toxicities. However, we don't know what it will do in combo. So starting at half the dose, pretty standard, and then if it looks well tolerated, we may then increase back up to the 150 Q6. But there could be, yes, some overlapping immune-related toxicities, so we want to just, you know, be cautious about that.
Yep. Talk a little bit about timing. You suggested maybe 40 patients. I mean, is this? Can we get some early insights next year, or is that too ambitious?
I think it's probably premature to-
Yep
T o say exactly, because I think importantly, we have to determine how does the combination look, how many dose levels are in the combination. But we do anticipate opening sites to get started this year and moving as quickly as we can. I think investigators are already on board and interested in this path because this, the MSS- CRC population really doesn't have any IO options. So I think this is something that investigators are paying attention to and will move quickly.
Yep, makes sense. There's also been a lot of activity and excitement, and progress in the CTLA space overall. Maybe you can just help us understand how you're thinking in terms of the competitive landscape here, and maybe compare it to some of the other molecules in development.
Yep, yep. Yeah, so I think, what's evolved, there's a few things that have evolved in the CTLA-4 landscape, you know, in recent years. I think, number one, we've learned something interesting about dosing. We've learned from data, from AstraZeneca on tremelimumab, that a single high dose of their anti-CTLA-4 actually had better efficacy than that same exposure broken into multiple doses. So that sort of tells us as a field that Cmax really drives efficacy for CTLA-4s. So that was important for us in dose selection. That's kind of a newer sort of insight. The other piece is that Fc-enhanced CTLA-4s that have added potency, but also the Fc enhancement, are demonstrating activity in colder tumors, including in MSS-CRC, but also other types. We believe that's due to T-reg depletion as a secondary mechanism, as well as the priming.
So we think that's playing a role. And then in terms of how we would differentiate, our whole sort of hypothesis is the more potency you're driving, which these Fc-enhanced CTLA-4s are highly potent, the more you're challenged with therapeutic index. So how do you get the Cmax you need, which is driving efficacy, and still have a tolerable drug, right? And I think for us, not seeing any grade four or five tox at all in our phase I at all doses was very encouraging. And then as we move to this 150 Q6 weeks, we feel like we have a really nice tolerability profile. So balancing that potency with tolerability and then now demonstrating that we can activate it selectively in the tumor, we think is how we would differentiate.
Then in terms of the trial design, the differentiation is the fact that we will be including those patients with liver metastases, which is about 75% of patients with colorectal, MSS colorectal cancer.
Yeah. So it sounds like you have all the pieces of the puzzle to kind of make this work together. Okay. Maybe just talk a little bit about your, the long-term opportunity for 101.
Yeah
I n terms of the ongoing collaboration with Roche, but just also broadly, your partnering strategy?
Yeah. So the way we're thinking about this opportunity with Roche, first of all, we're very obviously excited to partner with Roche. They really understand this space and the work that we're doing. In particular, they also know a lot about cytokines. And so they—I think they're very sophisticated in understanding these type of mechanisms. The MSS-CRC path that we're starting on, we think makes sense because of the efficiency and being able to do this, you know, without a randomized trial, right? So that makes a lot of sense, and being able to include the liver metastases in that population. But our first response was a non-small cell lung cancer, right?
So that would be another area we would like to potentially pursue with a partner, and be looking for the other opportunities to put this agent in combination. And I think those kinds of collaborations where we're both contributing make a lot of sense for us.
G otcha. Maybe we can shift gears to just 202. It's your second program. Maybe just give us a little bit of background on IL-2. Why is that an interesting target? Some of the challenges and how you can overcome those.
Yep. Yeah, so, IL-2 has had quite an up-and-down sort of rollercoaster ride. But the real, you know, the motivation why we keep chasing IL-2 is because we did see those long-term durable cures, right? Primarily in melanoma, but such a highly toxic molecule and really difficult to keep people on it long term. So we are focused. We have a beta gamma IL-2, tumor selective, and what we've seen, and very recently, as recently as last week, we've now cleared a dose level seven, which is well beyond our expectation for this molecule, which is a 4 mg per kg dose, very high dose. We're administering this completely in the outpatient setting. There are no signs of VLS, which is the key toxicity for IL-2. And we have patients now that are out, as far as 19 cycles, so over a year of therapy.
And that just speaks to the tolerability and the ability of a phase I population to stay on an agent like IL-2 with these doses for a year is pretty unusual. So we think this molecule is very well set up now to begin to look at what do we do next, and what are the combinations for this type of IL-2 that make sense? Checkpoints, T cell engagers, TIL therapies, for example. A lot of different mechanisms that could benefit from the combination with IL-2. And we've also been able to look at four different tumor biopsies from patients in the trial, and in all four of these, we saw increases in CD8s in the tumor and not in the periphery. So we're seeing that tumor selective activation. So I think we're on the right track.
Good.
And then we'll have to think about what is the right, you know, path forward for this molecule.
Just in terms of your choice of a beta-gamma IL-2, just maybe talk about that a little bit, and what's the latest thinking there, more broadly in the field? Is that the right way to go, or are there other ways to go?
Yeah. So, clearly historically, right, wild-type IL-2 has worked, right? That is sort of the original data. Our thesis is that when you have the alpha binding, right, that is the, the primary sort of, you know, where the majority of the binding will happen first with the molecule, and that really drives Tregs. So, we also have a program that we've developed that's alpha-biased IL-2 to stimulate Tregs, which would be an autoimmune, right? So IL-2 has this, like, yin and yang. Our take was that if we were gonna concentrate this molecule only in the tumor, we weren't worried about peripheral tox, we could just eliminate that, Treg stimulation, eliminate the alpha binding, go full beta gamma. I think that may be riskier with a systemic molecule because that's a little bit putting the brakes on the immune system.
Yep.
I don't think there's a complete right or wrong way, there are sort of flavors. But I think for us, with the tumor selectivity, that full beta gamma, avoiding Tregs, we thought made the most sense.
Yep.
That's the approach we've taken.
Got it. Can you just go back to the doses you've tested so with 202, and how you think about that going forward? I think you said you're at 7 mg.
Seventh dose level-
Seven, okay.
4 mg, 4 mg per kg, yeah.
Okay. I'm-
Yeah
G etting ahead of myself here.
Yeah. Yeah, so I think we are now, based on our preclinical data, sort of at the top of the dose-response curve, right? We don't anticipate, again, based on preclinical data, that going much beyond 4 mg per kg gets additional-
Yep
A dded benefit for efficacy, so I don't anticipate going much higher. We did open the phase II at a dose of 1.4 mg per kg, which we believed would be the bottom of the, you know, sort of efficacious response, and that has been enrolling. And now we're gonna open up a second dose level in the phase II, and we'll just be making a decision, do we do that at the 2.8, or do we do this at the new 4 mg per kg dose we just cleared? We will expand that top 4 mg per kg dose cohort in the phase I. A few more patients, make sure we really understand it before we make that decision. But I don't expect to go much beyond the 4 mg per kg dose.
In the phase II, it might be two doses, potentially, maybe three?
Yeah.
Okay.
I think most likely two.
Okay.
I think we'll pick a second dose and then go-
Okay
A nd then go.
When will we get sort of the next update from the phase I or the combo? And, and what should we expect? How many, y ou know, I know it's hard to answer sometimes-
Yeah
B ut, what's your general view on when you share data like-
Yes, yes. So I think for IL-2, we did announce that we do have a poster accepted at SITC, and we will be presenting the data we have to date for that molecule, which is going to be primarily phase I population, but we'll also have some phase II patients in that population. We do expect it to be at least 20 patients worth of data, treated at 1 mg per kg or higher. The fact that we continue to go higher, I think, is interesting, and we will wanna look to see, you know, what is that activity at these higher doses. We should have some of that data available in this dataset.
When you say activity, you mean responses on most patients, or maybe some patients may not be followed up long enough or just had a-
Yeah, I think it will be a mix. I think the higher doses, you know, like four, you know, are pretty recent, so we may not have very long-term, you know, follow-up and response data for some of these higher doses. But we'll look at everything we have to date, and especially focused on 1 mg per kg or higher for any anti-tumor activity.
Yep. And so the view here, any activity would be a positive, or is there a certain bar, or we just have to wait for the data mature to kind of think about?
Y eah, I think our view has been, you know, if, if we want to do, slightly better than what we've seen in this population, so this is, you know-
Yeah
H eavily pre-treated, you know, four to five prior lines of therapy, sort of all-comer population. Seeing something in the 10%-15% response rate for monotherapy IL-2 would be better than what's been shown in the past. And then with this safety and tolerability profile, you know, moving forward in combination, I think we'd like to see some evidence of monotherapy activity. But I think, you know, in this population, it may not be more than 10%-15%. That's kind of been our bar internally.
Yeah. So the view, just hit that bar with better tolerability, and then you have the combinability that gives you the boost, is that-
Yes.
Okay.
I think that's right.
Okay. All right. Maybe we can shift to 301. It's your third program. It's IL-12. Maybe just talk a little bit about that cytokine and why that could be interesting, and then maybe more broadly, some of the challenges that we've seen across the space.
Yep. Yeah, so our IL-12 molecule, we're very excited about, because when we think about IL-12, that mechanistically, historically, really has been uniquely able to turn a cold tumor warm, right? And to open up IO to all of those patients that have cold tumors who don't today benefit from IO. So that's really the hope. And historically, the challenge with IL-12, even more so than IL-2, even more so than CTLA-4, has been extreme toxicity including grade 5 tox, very early.
So, our preclinical data, you know, encouraged us, and allowed us to have a starting dose that's significantly higher than the MTD of, recombinant human IL-12. So it's, it's a high starting dose compared to what's been done historically, and that's because the masking. We've also seen, CRs preclinically after a single dose. So we see some impressive opportunity for this molecule to be very meaningful. We are in phase I dose escalation and expect to, by the end of this year, share safety data through the first three dose levels, and that begins to get us into the bottom of that meaningful, you know, dose response. And we think would be pretty important from a tolerability perspective if we can bring that molecule forward, you know, beyond dose level three.
So dose level three, you expect to see some activity, or is that kind of the low end, or we're kind of on the-
I think we're beginning to get-
Okay, yeah
I nto the lower end of where it becomes clinically meaningful exposures.
Yeah. So if you're kind of safe in those initial three, you can continue, continue to dose escalate, and you should be more in the effective range.
Yes.
Okay.
Yes.
Okay. Can you just talk about the platform between 202 and 303 in terms of the unmasking? In other words, how different is that? If you've de-risked sort of the unmasking-
Yep
O r you have confirmation of unmasking, with 202, does that sort of translate to 303? Or is there some reason why it may not?
Yeah. I think generally, because we're seeing... So all our approach overall, the MMP approach-
Yep
... is applied across all of the molecules. The structures are different, depending on if it's an antibody or cytokine or heterodimeric cytokine like IL-12. But the underlying sort of MMP-driven thesis is consistent, and we've built in redundancies in all of these molecules. So if one MMP is missing from a solid tumor, there are the presence of any other-
Yep
M ight, you know, then cleave the molecule. So I would say we feel pretty confident that what we're seeing with XTX101 and XTX202, in terms of the tumor-selective PD and the tumor-selective PK now should be expected to translate to IL-12. I would say there is some, you know, kind of readthrough to all of these.
Yeah.
Of course, all the molecules are different.
Right.
But it's not, you know, the structures are different.
Right.
But-
Steric issues , things like that-
Yeah
M aybe or something.
The design process we're applying-
Yeah
A nd now sort of confirming, right, is applied to each of these.
Do you use the same, well, the same MMPs, or I guess, do you use the same cleavage site across all three molecules, or is there?
Not necessarily.
Okay.
Because the, the 3D conformation of the linker and how that cleavage site is presented-
Okay
Can change things.
Okay.
So it's not a, y es, our CSO is in the back, so he can speak more to this after. But it's not a plug-and-play, so once we figure it out from one molecule, it—we still have to design each one. But the family of MMPs and literally over 1,000 human solid tumor samples we have in-house, we've studied the protease profiles. That work is applied, and then sometimes we use different, you know, linker designs and different cleavage sites, depending on the 3D conformation.
Gotcha. Sorry, I was referring to 303 before. I meant 301.
301.
The starting dose for 301, can you just compare that to where you started with 202?
202
2 01?
It's dramatically lower-
Yeah
B ut it's a much-
Because it's more potent.
M ore potent molecule.
Yes. Okay.
We're still starting about 10x beyond the MTD of human recombinant IL-12.
Yeah.
So it's a 5 mcg per kg starting dose, which seems small, but it's a pretty big dose of IL-12.
Yep. When you think about 301 and the potential opportunities in different tumor types-
Yep
I s it, is there overlap, or is it broader? Or how are you thinking about that at this point?
Yeah, I think what we would like to do, and it's very early to say-
Yep
Q hat we can do with IL-12, but I think what we would like to see is the ability to open up IO in more colder tumor types and then look at combinations. So we think if IL-12 works the way we hope... There may be some overlap, but it would be coming mechanistically f rom a different place, and also would potentially open up some tumor types that wouldn't otherwise be responsive to an IO and then start to combine.
So really, like getting that cold tumor microenvironment, you know, kind of turned on to a warm environment and then bring in additional IO mechanisms. And that's all about having the masking and that tolerability to be able to do that.
For 301, do you think you could see monotherapy activity or not? And do you think you need combination, or because it's more potent, assuming you can have the therapeutic window, that you may not need to do a combination?
Well, I think mechanistically... So a couple things. I think we know historically, IL-12, like intratumoral injection, it has shown antitumor activity on its own. Systemically delivered, I don't think we've done that safely yet, so we don't know. So I do think there is some opportunities based on historic data for IL-12 to have an antitumor effect on its own. But I think the benefit mechanistically is the synergy between, you know, the IL-12 mechanism, warming up that environment and then bringing in other things like checkpoint IL-2, for example.
Good.
So ultimately, I think the combination will be the way to go. But there have been historical, you know, antitumor effects with a monotherapy IL-12-
Yep.
Direct injection.
Yep, makes sense. Maybe just future sort of pipeline or platform, and how you're thinking about that. Do you-- I guess, what's next, or right now is the focus just get these three molecules sort of through a certain hurdle, and then you can kinda think about pipeline expansion?
We already have some pipeline expansion.
Okay. Yeah.
So one of the things that I think we're very excited about, but we had to make sure we had all the parts really well understood, is what we would call multifunctional molecules, 'cause we talked a lot about combinations. If you can put, sort of complementary mechanisms together in the same molecule and then bring them together directly to the right cellular target, you get dramatic synergy. And one of the areas that we're very excited about is a PD-1/IL-2 multifunctional molecule that we have. There's been some prior work in this space. Roche has done some great work in this space, but it's been really limited by extreme toxicity, right? Because now you're adding two very, very potent molecules.
The tox is very challenging. So we have a tumor-selective PD-1 IL-2 that allows us to get that, synergistic activity, not sort of passively, by just, you know, combining these two and hoping the concentrations come to the right place, but really delivering the IL-2 with the PD-1 directly to the right cells. It's dramatically different in vivo and preclinically. I'm very excited about that.
Now that we understand our masking-
Yep.
Works, we know how to apply it to cytokines, we know how to apply it to antibodies. These multifunctional molecules, I think, will be a big part of the future.
Yeah. Interesting. When, when can we sort of push that into that?
Yes. Yes. I think, I think it's showing up in our pipeline now, which-
Okay
Y ou know, we're getting there. We'll be looking to, you know, identify really that development candidate, the right design.
Yep.
We put a lot of constructs into the non-human primate studies before we choose the final, but I think we're getting near that point and, you know, would like to move toward this being part of our, you know, our IND enabling work soon.
Yep. And maybe you can just talk about the other, the other side of the coin here, which is funding all these sort of programs and keeping them moving forward. Maybe just talk about cash and your runway currently.
Yes. Yes. So we have runway into the second quarter of next year, through the second quarter of next year. And we're running a lot of programs. So I think one of the opportunities for companies like ours when we have a lot to pursue, is exactly the type of collaboration we just announced with Roche. Where they're also contributing, you know, from a perspective not just of the supply agreement, but also-
Yep
C ontributing to the development and to the, investment in that development. Those are the kinds of partnerships I think we'll be continuing to look for, because there are so many opportunities where to go with these molecules.
Yep. Maybe just the last question to wrap up. You know, we talked, we touched on this already, but just next catalysts, kind of from now until mid-next year, can you just map those out for us quickly?
Sure. Yeah. So toward the end of this year, we will have our complete data set to date that we have for IL-2. So that'll be some of the phase I and some phase II patients treated at 1 mg per kg or higher. Expect to have about 20 patients in that data set. That'll be November. End of year, we'll look at IL-12 safety through dose level 3, and then we also anticipate initiating that combination study with the atezo and opening that trial up toward the end of this year.
Okay, great. So 10 seconds left. I think we can stop there. And, thanks, René.
Thank you.
Great to catch up with you.
Likewise. Thank you.